Relevance and limitations of these findings

We have begun to characterize a novel rodent model of birth asphyxia in collaboration with Prof Kai Kaila’s group at the University of Helsinki, who originally developed this paradigm. This model encompasses many of the pathological features of in utero asphyxia, including clinically relevant O2, CO2 and Trec ranges. It resulted in acute inflammation in certain well-defined areas of the brain without overt neuronal death or systemic injury. When these animals were followed into adulthood, they showed a phenotype of anxiety and impulsivity, without neuromotor deficiencies.

The combination of anxiety and impulsivity is a characteristic of attention-deficit/hyperactivity disorder (ADHD).²⁰¹ ADHD is a multifactorial developmental disorder which usually presents itself in school-age children with an overall prevalence of approximately 2-3%.²⁰¹ While its pathogenesis is largely unknown, epidemiological studies suggest that genetic as well as pre- and postnatal environmental factors are involved.²⁰¹ Recent studies indicated perinatal hypoxia as a potential risk factor in the development of ADHD.²⁰²

As discussed in the Introduction, approximately 10% of all newborns require some form of assistance at birth and 1% need vigorous resuscitation, while only a small subset of these children will go on to develop neonatal encephalopathy.⁹ Thus, there is a large cohort of children who likely suffer some above-normal level of asphyxia at birth,

73

but will not require intensive care and thus will not receive neuroprotective interventions such as hypothermia. It has been suggested long ago that if birth asphyxia is a continuous spectrum, then infants who die or suffer moderate to severe HIE must only represent one end of the spectrum and there should be a large group of neonates, who suffered relatively mild asphyxia.²⁰³ This group of infants is not seen in asphyxia trials and their first interaction with healthcare might be much later, when neurodevelopmental problems emerge. Recently there have been attempts to follow such large cohorts into childhood to prospectively determine the effects of perinatal factors. These cohort studies have employed stratification criteria based on Apgar-scores,^13,14 a need for resuscitation,^15,16 or acidosis at birth.¹⁷ Their results have been equivocal, a few studies providing some level of confirmation to the theory of a continuous birth asphyxia spectrum,^14,15 but the effect size of mild asphyxia appears to be rather low and of questionable clinical significance. However, we should note that these stratification criteria, such as the 5-minute Apgar score are generally poor predictors of outcome even in children with moderate to severe HIE, hence their usefulness in more complex and long-term neurodevelopmental disorders are probably rather limited. Unfortunately, we currently have no better means for designing such cohort studies and for stratifying patients. Hence the potential causative role of perinatal hypoxia probably has to be first elucidated in well-controlled animal studies. We propose that this novel rodent model of birth asphyxia originally developed by Prof. Kai Kaila’s group might be optimally suited for such work. It offers the possibility to identify neuroanatomical structures and functional connections specifically affected by mild perinatal asphyxia, the role of which could in turn be confirmed in human ADHD neuropathology. Even if these abnormalities can only be identified in later childhood, it would still offer the possibility to study the role of neuroprotective interventions on these deficiencies. These therapies could theoretically be applied around birth or possibly later on. However, without the means of definitive diagnosis at the time of intervention, these therapies will have to be designed as preventive measures, which benefit those in need of them, but does not impose a risk to others. One such therapy might be the restoration of normocapnia after asphyxia in a graded, instead of an abrupt manner.¹⁶⁶

Certain limitations to our study should be noted here. The rodent brain has an

74

inherently different structure than the human CNS. Additionally, as discussed in the Introduction, the optimal postnatal age for developmental comparison is unclear. We have chosen P7 animals largely to obtain results comparable to the most widely-used rodent models and to be able to identify model- and not age-specific differences. There are convincing arguments that a P10-12 rat might be more comparable to the term human newborn.²⁰⁴ Another important difference from human perinatal care is the lack of resuscitative efforts in this model. While our goal was to use minimal intervention, and intensive care would have required a number of significantly more stressful actions, there are other groups who are exploring these opportunities for translational preclinical studies with good results.²⁰⁵ Finally, extrapolating from rodent behavioral studies to humans is of course never straightforward, therefore our findings in this model will have to be confirmed in human patients. This is both the most challenging and the most promising aspect of our future work.

In summary, we have begun to characterize a novel, non-invasive rodent model of mild perinatal asphyxia. The animals surviving this insult showed localized inflammation in the brain without overt neuronal death, axonal injury or white matter damage. In adulthood, these animals displayed increased anxiety and motor impulsivity.

This correlates well with the human phenotype of ADHD. These findings support previous epidemiological data suggesting a correlation between mild perinatal hypoxic events and neurodevelopmental impairments in later childhood, such as ADHD. This novel preclinical model might be instrumental in the future investigation of ADHD and childhood neurodevelopmental deficiencies.

75